Constructing a Community : Myths and Realities of the Open Development Model

(1)

University of Helsinki, Institute of Behavioural Sciences Studies in Educational Sciences 242

Stephanie Freeman

CONSTRUCTING A COMMUNITY

Myths and Realities of the Open Development Model

Academic dissertation to be publicly discussed, by due permission of the Faculty of Behavioural Sciences at the University of Helsinki

in Lecture room 302 (Athena, Siltavuorenpenger 3 A) on the 17^th of December, 2011 at 12 o’clock.

Helsinki 2011

(2)

(3)

Custos

Professor Reijo Miettinen, University of Helsinki Supervisor

Professor Reijo Miettinen, University of Helsinki Pre-examiners

Docent, PhD Jussi Silvonen, University of Eastern Finland

Assistant professor Yuri Takhteyev, University of Toronto, Canada Opponent

Docent, PhD Tere Vadén, University of Tampere, Finland

Cover illustration Robert Freeman

Unigrafia, Helsinki

ISBN 978-952-10-6993-2 (pbk) ISBN 978-952-10-6994-9 (PDF) ISSN-L 1798-8322

ISSN 1798-8322

(4)

University of Helsinki, Institute of Behavioural Sciences Studies in Educational Sciences 242

Stephanie Freeman

Constructing a Community

Myths and Realities of the Open Development Model Abstract

The open development model of software production has been characterized as the future model of knowledge production and distributed work. “Open development model” refers to publicly available source code ensured by an open source license, and the extensive and varied distributed participation of volunteers enabled by the Internet. Contemporary spokesmen of open source communities and academics view open source development as a new form of volunteer work activity characterized by “hacker ethic” and “bazaar governance”.

The development of the Linux operating system is perhaps the best know example of such an open source project. It started as an effort by a user- developer and grew quickly into a large project with hundreds of user-developer as contributors. However, in “hybrids”, in which firms participate in open source projects oriented towards end-users, it seems that most users do not write code. The OpenOffice.org project, initiated by Sun Microsystems, in this study represents such a project. In addition, the Finnish public sector ICT decision- making concerning open source use is studied. The purpose is to explore the assumptions, theories and myths related to the open development model by analysing the discursive construction of the OpenOffice.org community: its developers, users and management.

The qualitative study aims at shedding light on the dynamics and challenges of community construction and maintenance, and related power relations in hybrid open source, by asking two main research questions: How is the structure and membership constellation of the community, specifically the relation between developers and users linguistically constructed in hybrid open development? What characterizes Internet-mediated “virtual” communities and how can they be defined? How do they differ from hierarchical forms of knowledge production on one hand and from traditional volunteer communities on the other?

The study utilizes sociological, psychological and anthropological concepts of “community” for understanding the connection between the “real” and the

“imaginary” in so-called “virtual” open source communities. Intermediary methodological and analytical concepts are borrowed from discourse and

(5)

rhetorical theories. A discursive-rhetorical approach is offered as a methodological toolkit for studying texts and writing in Internet communities.

The empirical chapters approach the problem of “community” and its membership from four complementary points of views. The data comprises mailing list discussion, personal interviews, web page writings, email exchanges, field notes and other historical documents. The four viewpoints are: 1) the community as conceived by volunteers 2) the individual contributor’s attachment to the project 3) public sector organizations as users of open source 4) the community as articulated by the Community Manager.

I arrive at four conclusions concerning my empirical studies (1–4) and two general conclusions (5–6). 1) Sun Microsystems and OpenOffice.org Group- ware volunteers failed in developing necessary and sufficient open code and open dialogue to ensure collaboration thus splitting the “Groupware community” into volunteers “we” and the firm “them”. 2) Instead of separating intrinsic and extrinsic motivations, I find that volunteers’ unique patterns of motivations are tied to changing objects and personal histories prior and during participation in the OpenOffice.org Lingucomponent project. Rather than seeing volunteers as a unified community, they can be better understood as “independent entre- preneurs” in search of a “collaborative community”. The boundaries between work and hobby are blurred and shifting, thus questioning the usefulness of the concept of “volunteer”. 3) The public sector ICT discourse portrays a dilemma and tension between the freedom to choose, use and develop one’s desktop in the spirit of open source on one hand and the striving for better desktop control and maintenance by IT staff and user advocates, on the other. The link between the global OpenOffice.org community and the local end-user practices are weak and mediated by the problematic IT staff-(end)user relationship. 4) Authoring community can be seen as a new hybrid open source community-type of mana- gerial practice. The ambiguous concept of community is a powerful strategic tool for orienting towards multiple real and imaginary audiences as evidenced in the global membership rhetoric. 5) The changing and contradictory discourses of this study show a change in the conceptual system and developer-user relationship of the open development model. This change is characterized as a movement from hacker ethic and bazaar governance to more professionally and strategically regulated community. 6) Community is simultaneously real and imagined, and can be characterized as a “runaway community”. Discursive- action can be seen as a specific type of online open source engagement. Hierar- chies and structures are created through discursive acts.

Key words: open source software, open development model, community, motivation, discourse, rhetoric, developer, user, end-user

(6)

Helsingin yliopiston käyttäytymistieteiden laitos Kasvatustieteellisiä tutkimuksia 242

Stephanie Freeman

Yhteisöä rakentamassa

Avoimen kehittämismallin myytit ja todellisuudet Tiivistelmä (Abstract in Finnish)

Avoin kehittämismalli on tullut yleiseen tietoisuuteen Suomalaisen Linus Tor- valdsin 1990-luvun alussa käynnistämän Linux-käyttöjärjestelmän kehittämis- projektin kautta. Avoin kehittämismalli viittaakin Internetin välityksellä tapah- tuvaan, maantieteellisesti hajautuneeseen maailmanlaajuiseen ohjelmistojen kehittämistapaan, jossa kehittämisen kannalta olennainen lähdekoodi on julki- sesti saatavilla. Malli perustuu kehittäjien ja virheentunnistajien vapaaehtoiseen osallistumiseen ja tehtävien valintaan. Kaupalliset toimijat kiinnostuivat 2000- luvun alussa avoimen kehittämismallin hyödyntämisestä, ja niin syntyi loppu- käyttäjälle suunnattuja “hybridiprojekteja”, joissa suurin osa käyttäjistä ei kuitenkaan kirjoita koodia. Ilmiö ei siis ole enää pelkästään hakkereiden harras- telemista vaan yhä useammin elimellinen osa yrityksen liiketoimintaa. Tähän uuteen hybridiin yhteisön ja yrityksen toimintamuotoon liittyy haasteita, joihin tämä poikkitieteellinen väitöstutkimus pyrkii vastaamaamaan.

Tutkimuksen kohteena on Sun Microsystems -yrityksen käynnistämä avoimen lähdekoodin toimisto-ohjelmistoa kehittävä projekti OpenOffice.org.

Lisäksi tutkimuksessa tarkastellaan Suomen julkista sektoria avoimen lähde- koodin hyödyntäjänä. Väitöskirjan tarkoituksena on tutkia avointa kehittämis- mallia ja sen yhteisöjä koskevia olettamuksia, teorioita ja myyttejä analysoimal- la OpenOffice.org kehittäjien motivaatiota, käyttäjiä ja johtamista.

Laadullinen tutkimus tuottaa tietoa hybridin avoimen lähdekoodin Internet- yhteisön rakentamisesta ja ylläpidosta sekä näihin kietoutuvista valtasuhteista.

Tutkimuksen pääkysymykset ovat seuraavat: 1) Miten avoimen kehittämismal- lin muutos vaikuttaa yhteisön rakenteisiin ja jäsenkunnan koostumukseen ja erityisesti kehittäjien, tekijöiden ja käyttäjien keskinäisiin suhteisiin? 2) Mikä on ominaista Internet-välitteiselle “virtuaaliyhteisölle” ja miten se voidaan määritellä ja miten se eroaa hierarkkisista tuotannon organisointimuodoista ja toisaalta perinteisistä yhteisöistä?

Tutkimuksen teoreettisessa viitekehyksessä hyödynnetään sosiologisen, ant- ropologisen ja psykologisen tutkimuksen yhteisökäsityksiä, joiden avulla pyri- tään ymmärtämään todellisen ja kuvitteellisen yhteyttä virtuaaliyhteisöissä.

Metodologisia ja analyyttisiä käsitteitä ammennetaan diskurssianalyysin ja

(7)

retorisen analyysin teorioista. Samalla kehitetään diskursiivis-retorinen metodo- loginen välineistö tekstien ja kirjoittamisen tutkimiseen Internet-yhteisöissä.

Empiirisissä luvuissa yhteisöä koskevaa puhetta ja yhteisön määrittelyä lähestytään neljästä toisiaan täydentävästä näkökulmasta:1) yhteisö vapaaehtoisten määrittelemänä, 2) vapaaehtoisten kiinnittyminen yhteisöön, 3) Suomen julkinen sektori avoimen lähdekoodin hyödyntäjänä, 4) yhteisö yhteisöjohtajan määrittelemänä. Aineisto koostuu sähköpostituslistojen keskusteluista, henkilö- kohtaisista puhelinhaastatteluista, web-sivujen kirjoituksista, blogeista, tutki- mushenkilöiden kanssa käydystä sähköpostikirjeenvaihdosta, kenttämuistiinpa- noista sekä historiallisista dokumenteista.

Empiiristen analyysien pohjalta muotoillaan neljä lukukohtaista johtopäätös- tä (1–4) ja kaksi yleistä johtopäätöstä (5–6). 1) Yrityksen ja vapaaehtoisten yhteistyö OpenOffice.org Groupware- projektissa epäonnistui. Osapuolet eivät kyenneet muodostamaan riittävää dialogia. Tämän sekä yrityksen päätöksenteon avoimuuden kyseenalaistamisen seurauksena Groupware-yhteisö jakaantui vapaaehtoisiin ”meihin” ja yritykseen ”heihin”. 2) Sisäisten ja ulkoisten moti- vaatioiden jaottelun sijaan löytyy yksilöllisiä muuttuvia motivaatio-yhdistelmiä, jotka ovat sidoksissa vapaaehtoisten ihmissuhteisiin ja heidän muuttuviin tekno- logisiin objekteihin ennen kieliteknologia- projektiin osallistumista ja sen aikana. Harrastuksen ja työn raja on hämärtynyt, minkä perusteella vapaaehtoi- suuden-käsite voidaan kyseenalaistaa. 3) Julkisen sektorin IT-diskurssi näyttäy- tyy dilemmaattisena, toisaalta puheena käyttäjän valinnanvapauden takaamisen tärkeydestä ja toisaalta puheena käyttäjän työpöytäohjelmistojen paremmasta kontrolloitavuudesta. OpenOffice.org yhteisön ja paikallisen käyttäjän välinen yhteys on heikko ja sitä välittää ongelmallinen IT-henkilöstö-käyttäjä-suhde.

4) Yhteisön käsikirjoittaminen voidaan nähdä uutena hybridin yhteisön johta- miskäytäntönä. Yhteisö-käsite on monimerkityksinen, ja siksi strategiseksi välineeksi sopiva. Sitä voidaan käyttää moninaisten todellisten ja kuviteltujen yleisöjen vetoamiseen, kuten yhteisöjohtajan uusi globaali jäsenyysretoriikka todistaa. 5) OpenOffice.org-projektin ja julkisen sektorin vaihtuvat ja ristiriitai- set diskurssit havainnollistavat avoimen kehittämismallin muuttuvaa kehittäjä- käyttäjä-suhdetta. Tätä kehitystä luonnehtii muutos hakkeri-etiikkaa ja basaari- hallintoa korostavasta puheesta strategisesti säädellympään yhteisöpuheeseen.

6) Yhteisönä OpenOffice.org on samanaikaisesti todellinen ja kuvitteellinen, ja sitä voidaan kuvata “pakenevaksi yhteisöksi”. Diskursiiviset teot ovat luonteen- omaisia hybridiin projektiin osallistumisessa. Hierarkiat ja rakenteet syntyvät diskursiivisten tekojen kautta.

Avainsanat: avoin lähdekoodi, avoin kehittämismalli, yhteisö, motivaatio, diskurssi, retoriikka, kehittäjä, käyttäjä, loppukäyttäjä

(8)

ACKNOWLEDGEMENTS

While this research process was characterized by long expanses of lonely searching, reading, ideation and writing–essentially an effort after personal meaning–this book would not be here if it weren’t for the communities that have influenced my thinking. During these seven years I have had the privilege of meeting inspiring people both inside and outside academia thus engaging in multiple conversational realities.

I want to start by thanking my two external reviewers, Docent Jussi Silvonen from the University of Eastern Finland, and Assistant Professor Yuri Takhteyev from the University of Toronto (Canada), for your critical and constructive comments that gave me the opportunity to polish up parts of the dissertation and take into account recent events in the turbulent world of open source in the form of an Epilogue.

I am deeply grateful to my supervisor Professor Reijo Miettinen for seeing the potential in me in the first place, and inviting me to join your research group

“Innovations and Organization of Research Work”. I find your passionate and sincere attitude towards science truly inspiring. I am also immensely grateful for our discussions and your feedback on my manuscripts. I am glad we have not always agreed on issues, as our debates have made me stronger in argumentation and more independent as a researcher. Thank you for giving me the intellectual freedom to pursue my interests and intuitions. We have also shared many laughs, and enjoyed discussions on birds and other animals in general.

Thank you for also sharing your knowledge in these domains.

I was privileged to work at the Center for Research on Activity, Develop- ment and Learning (CRADLE), University of Helsinki, lead by Professor Yrjö Engeström. Thank you for providing the infrastructure necessary for carrying out my research. I was also given the opportunity to participate in the unique and intellectually stimulating doctoral program developed and conducted by Professors Yrjö Engeström, Reijo Miettinen and Jaakko Virkkunen. Thank you for the excellent training. I am also grateful to Kari Toikka for his teaching on the works of Karl Marx as part of the doctoral program.

I have had a joyful time with my 2003 doctoral school “buddies” Kirsi Kal- lio, Annarita Koli, Ulla-Maaria Engeström, Auli Pasanen, Anna Rainio, Marika Schaupp, Juha Siltala and Marianne Teräs. What a vibrant and creative class we were! Thank you for your friendship. I also want to thank former and present members of the Center for Activity, Development and Learning (CRADLE) work community for making it my home for so many years. Specifically, I owe my gratitude to former and present members of Professor Reijo Miettinen’s

(9)

research group: Janne Lehenkari, Juha Leminen and Juha Tuunainen for collaboration, and Tarja Knuuttila and Erika Mansnerus for the many inspiring discussions on science and technology studies. I also want to express my gratitude to Olivier Irrmann, Heli Kaatrakoski, Anu Kajamaa, Leena Käyhkö, Teija Mankkinen and Jenny Vainio for our lively “corridor discussions”. Special thanks go to those who directly commented on parts of the introductory chapters: Hannele Kerosuo and Kari Toikka from CRADLE, and Professor Sampsa Hyysalo, Mikael Johnson, Jouni Juntunen and Samuli Mäkinen from the research group “Innovations, Users and Communities” at Aalto University to which I currently contribute. I also want to thank Annalisa Sannino for our discussion on “idelogical dilemmas”.

I gratefully acknowledge the funding granted to me by the Finnish Post- Graduate School in Science, Technology and Innovation Studies (Tieteen, teknologian ja innovaatioiden tutkimuksen valtakunnallinen tutkijakoulu, TiTeKo), Emil Aaltonen Foundation (Nuoren tutkijan apuraha) and the Univer- sity of Helsinki.

The Finnish Post-Graduate School in Science, Technology and Innovation Studies was lead by Professors Reijo Miettinen and Marja Häyrinen-Alestalo.

I am truly grateful for the opportunity to participate in the versatile and well- organized set of seminars in the interdisciplinary field of science and technology studies. I am also indebted to Nina Honkela, a classmate in TiTeKo, for socializing me into the Assistant Editorship of the journal Science and Technol- ogy Studies.

I am grateful to Outi Grotenfelt, whom I met by coincidence in 2007 when I started searching for more tangible forms of open source in the Finnish public sector. Thank you for your companionship in the data collection.

I would also like to express my gratitude to Michael Freeman who took care that my language was flawless. I appreciate it that you see language as being simultaneously about form and content. I am also grateful to Tuomo Aalto for preparing this book for print.

I owe my deep gratitude to all OpenOffice.org volunteers and employees, and those members of the Finnish public sector IT staff who kindly agreed to be interviewed. Without your participation and openness this research would not have been possible.

My life would not be complete without music, drums and my band. Thank you A-Ilo (A-Joy) for the carefree moments that we have spent in the superb Nöykkiö studio over the past four years. Your friendship and encouragement have been of outmost importance to my wellbeing.

I also acknowledge the importance of my yoga and meditation practices and communities to making it through to this point. Since scientific practices are essentially about conceptual thinking, it is not uncommon that the brain contin-

(10)

ues to work in this mode “after work”. However, for me, this is not the best way to be creative. The brain needs a different kind of relaxed state where thoughts can be merely observed. My thanks to you, my wonderful yoga teacher training buddies for our explorations on the “self” and the “I”.

My close friends have seen my ups and downs and have always been there for me. For that I am tremendously grateful. My caring and wise parents Anneli and Michael Freeman, and my brother and close friend Robert Freeman, have offered me their unconditional love and support in pursuing my academic career. Thank you for the inspiring discussions and psychological support.

I dedicate this book to my beloved husband Joni Freeman. You were the fount and inspiration of this research.

Helsinki, 27^th of September 2011 Stephanie Freeman

(11)

(12)

1 INTRODUCTION

...I feel a little shy rhapsodizing about "community," especially as I live in California and feel, justifiably, that the term, community, has lost through overuse the precision of its contours and gained in weak ex- change a smooth but uselessly warm California feel. But would "collaborative environment" really do instead? No: For Open Source, the term

"community" is apt; it just should be used precisely, and indicate a working community, predicated on trust and collaboration...

As the quotation above by the OpenOffice.org Community Manager on the project’s website indicates, “community” is not an easy concept to use when trying to characterize open source software development, even for a community member. However, in everyday speech, as well as in academic writings about open source, values such as togetherness, connectedness, collaboration, shared goals and motives, socialization, and collective ownership are often mentioned.

Too little thought is given to its many uses and rhetorical power. Likewise the concept of open source and the open development model associated with it have been characterized in academic writings and in open source advocates’ speech as ideally empowering the user with respect to access, choice, development and distribution of software tools (e.g. Benkler, 2006, Lessig, 1996; von Hippel, 2005). Contemporary spokesman and scholars view open source development as a new form of volunteer work activity characterized by “hacker ethic” (Hi- manen, 2001) and “bazaar governance” via the self-selection of tasks by volunteers (Raymond, 2000). Because the open development model is conceived as the information society’s future model of globally distributed and shared work (Moon & Sproull, 2002) and knowledge production (Benkler, 2003), understanding its developmental dynamics is necessary. Further, the question of intellectual property rights and the way software production should be organized is a central issue in the discussion of the development of the “information society” (e.g. Boyle, 1996).

While open source projects may to some extent live up to these communita- rian ideals, a critical historically rooted viewpoint is missing. For instance, people involved in the development of the paradigmatic open source project

“Linux” used to be programmers with a personal need for a working operating system. Its development started as an effort by a user-developer and grew quickly into a large project with hundreds of user-developers as contributors. As the project grew it had to be modularized, which meant breaking up the code

(19)

2

into manageable units. Despite this, the division of labour did not change:

competent user-developers continued to do the coding.

However, with the emergence of “hybrid”¹ firm-initiated and sponsored open source projects oriented towards end-users, it seems that the people doing the coding are not necessarily the people doing the using. The “OpenOffice.org”

project studied here, represented such a project. It developed an end-user office application also named “OpenOffice.org”. The change in the object of activity of the open source community from programmer software to end-user software inevitably changes the nature of that community, the developer-user relationship, and the discourse of community and membership constellation. In the OpenOffice.org project a set of office software applications were developed in which both volunteers and the firm’s employees were working together. Since volunteer members can leave the project at any time, hybrid projects are faced with the challenge of attracting new members (especially volunteer programmers) and retaining old ones. While participation is voluntary, community management and construction in open source hybrids is also intentional.

This study aims at shedding light on issues of community construction and maintenance in hybrid open source by asking the questions 1) How is the structure and membership constellation of the community, specifically the relation between developers and users, discursively constructed in hybrid open development? 2) What characterizes Internet-mediated “virtual” communities and how can they be defined? How do they differ from hierarchical forms of knowledge production on the one hand, and traditional volunteer communities on the other? Today firms are to a growing extent utilizing or experimenting with open Internet-enabled platforms of product development, interacting with “lead- users” (von Hippel, 2005) or “prosumer communities” (Tapscott & Williams, 2007) or “crowdsourcing” (Howe, 2008). Hence, the findings could be also of value for companies acting in volunteer communities.

This study seeks to neither debunk nor romanticize open source communities. However, in order to go beyond “Free/Libre Open source” (FLOSS)² rhetoric and grasp the lived reality, FLOSS technology too should be approached as an inherently power-laden and ambivalent phenomenon (Berry &

Moss, 2007; Winner, 1985; Woolgar & Cooper, 1999). ICT technology and text are inherently intertwined (Woolgar & Cooper, 1999, p. 443), which suggests that technology has no meaning without the human discursive practices associated with its development and use. Hence it is important to explore the rela-

1 The metaphor “hybrid” is borrowed from biology and it designates “an offspring resulting from cross-breeding” (http://en.wikipedia.org/wiki/Hybrid) not capable of reproducing.

2 “FLOSS” is a term used by researchers for acknowledging both the ideological and pragmatic dimensions of free and open source software.

(20)

3 tion between academic discourses and actual discursive FLOSS practices. The purpose of this study is to explore the assumptions, theories and myths related to the open development model by analysing the discursive construction of the OpenOffice.org community: its developers, users and management, and Finnish Public Sector Information and Communication Technology (ICT) decision making concerning FLOSS use.

The theoretical approach used in this study combines sociological, psychological and anthropological concepts of community. These are used as complementary viewpoints for understanding community and for analysing the relation between the “real” and the “imagined” in so-called “virtual” FLOSS communities. Since writing is the dominant form of communication in the studied online community, an approach that focuses on language use in the construction and emergence of the community is justified. Hence, intermediary methodological and analytical concepts are adopted from discourse theories. I briefly outline the discursive-rhetorical approach (Billig et al. 1996; Fairclough, 1992; Mulhauser

& Harré; 1999; Shotter, 1993; 2003; Skinner, 2006) further developed in this publication as a methodological toolkit for studying Internet communities.

“Virtual or online ethnographic” data such as mailing list discussions and web-pages (Hine, 2001; 2008, pp. 257–271) were complemented with phone interviews between 2003 and 2007 from the large hybrid commercially sponsored open source software development project “OpenOffice.org”. This period was important because the text editor OpenOffice.org was going through significant changes and the project seemed to be growing fast. Additional face-to-face interview data were collected in 2007 from four end-users organization from the Finnish Public Sector, who either used or had considered using FLOSS-based tools. Combining these sites enabled the exploration of the relation between global development and local use.

The notion of community is approached from four complementary points of view along with complementary methodological tools, levels of analysis and sets of data:

Volunteers as members of the community. I analyse volunteers' conceptions of the hybrid community by analysing the themes that emerged from discussions among volunteers and between volunteers and the firm and the use of the linguistic device of contra-positioning in an episode leading to the demise of the OpenOffice.org sub-project Groupware.

1. The individual contributor’s attachment to the project. OpenOffice.org Lingucomponent volunteers’ changing motivations are analysed through types of contributions and personal paths of participation.

(21)

4

2. The end user as a member of the community. I analyse public sector end-users' argumentation for and against open source tools by analysing dilemmatic discourses.

3. The Community Manager as a member of the community. I analyse the OpenOffice.org Community Manager’s articulation of the community on the project's web pages during 2000-2007 through changing community membership categories.

The book is organized as follows. Chapter 2 describes the historical context this study and foregrounds the research task and research questions. Chapter 3 introduces and evaluates evolving conceptualizations of open source. Chapter 4 outlines the theoretical and methodological approach. It comprises two distinct sections. The first part introduces the four community concepts used as sensitiz- ing resources. Then it describes a methodological and analytical toolkit based on a discursive-rhetorical approach. Chapter 5 describes the field research process and data-collection. The empirical chapters 6, 7, 8 and 9 address specific research questions which can be seen as distinct entities and fields of research in their own right. Each empirical chapter includes a discussion of its findings.

Chapter 10 draws together the empirical conclusions and then discusses the main conclusions and their generalizability. Issues of validity and reliability are addressed throughout the writing process (Cohen and Manion, 2000).

(22)

5

2 THE HISTORICAL CONTEXT FOR STUDYING OPENOFFICE.ORG

One of the most pressing issues in discussions concerning the development of the information society has been how to organize the production of software (e.g. Boyle, 1996). Software production can be distinguished roughly by two different kinds of business logics. The proprietary model relies on the private ownership of software code and charges for software licenses. The open source model relies on the public ownership of code and searches for income in additional services, application and hardware development. However, when firms started joining open source communities in the turn of the millennium, hybrid communities started to emerge. In addition, a firm may today conduct both open and proprietary development and consultancy. Open source is indeed increasingly used as part of the firm’s services. This trend has followed to a greater extent from new companies building their solution on open source rather than established firms changing their products or business models.³ Hence, the open-closed dichotomy is not so straightforward in today’s business environment. Further, also the “open” model of software production is characterized by different degrees of openness. While the Free Software Foundation advocates

“free software” as an ideology (Stallman, 1984), the Open Source Initiative promotes “open source software” as a pragmatic approach to producing better software and as a viable alternative to proprietary software⁴.

In this study I define the open development model (ODM) as Internet- mediated geographically dispersed software development activity, in which the source code is publicly accessible, modifiable and redistributable to programmers. The starting point and necessary condition for this kind of activity has been volunteer participation and self-selection of tasks (Weber, 2004, p. 62). I prefer to use the notion “open development model” and not “free and open source software” because it directs attention to the organization, community, and the user-developer relationship thus making it easier to explore the hybrid model of open source, which is the focus of my analysis. Hence, the purpose of this chapter is to provide background and motive for the research task (see Figure 1 in chapter 2.4) by rendering visible the changing user-developer- relationship in the open development model.

3 See Rönkkö et.al. (2009).

4 See http://www.opensource.org/.

(23)

6

2.1 Promises of the open development model: user free- dom, democracy and transparency⁵

It is claimed that open source can democratize innovation, empower the user, and make Internet-based society more transparent (e.g., Benkler, 2003; 2006;

Lessig, 1999; 2009; von Hippel & Krogh, 2003; von Hippel, 2005). von Hippel (2005, p. 5) uses open source as an exemplar of user-driven innovation and anticipates a future where a “few fits all approach” no longer appeals to heterogeneous user needs; where users are increasingly able to innovate for them- selves; and where public policy making should support user innovation. Franke

& von Hippel (2003) propose that in order to better respond to heterogeneous market demands users be equipped with “innovation toolkits” or “user toolkits”.

Such tools are already used in the open source Apache project. They claim that a user-driven approach can better satisfy an array of user needs. Franke & von Hippel (2003) suggests that users have the kind of “sticky information” that is costly to acquire, transfer, and use in a new location. Thus if users are innovators, the information needed in technical decision-making is readily there.

Moreover, innovations developed by users have the potential to benefit non- innovators too (Franke & von Hippel, 2003, p. 1200). von Hippel’s view of future innovative activity envisions firms externalizing the development of new products and services to these “innovation communities”. Henceforth, he claims, firm strategies will be based on the utilization of the ideas and proto- types created by these communities.

Benkler (2003; 2006) regards users’ freedom to choose and join projects without requesting anyone’s permission the foundation of peer production. He cites open source development as an exemplar of what he calls “commons- based peer-production”. Benkler underlines that socio-technical systems of peer production offer not only an important medium of production for various kinds of information goods, but also serve as a context for positive character formation (Benkler, 2006, pp. 394–395) by fostering important moral and political virtues such as democracy, autonomy and social justice (p. 419; Berry & Moss, 2007; Perry and Fitzgerald, 2005).

Lawrence Lessig (1999) takes a transparency perspective on open source in his book Code and Other Laws of Cyberspace. He sees that open source could be a solution for making cyberspace more transparent and hence more demo- cratic. He maintains that cyberspace is a different space from the real physical place we inhabit. Lessig sees software architecture as a kind of law in the sense that it controls what people can and cannot do. In my mind, this code-as-law or transparency or control-approach to technology resonates with Langdon Win-

5 This section comprises some slightly modified paragraphs of a chapter that will be published in an article in the Information Technology and People journal (see Freeman, forthcoming).

(24)

7 ner’s (1985) view of technology as inherently political. Woolgar & Cooper (1999) urge science and technology researchers to explore the essential ambivalence of artifacts:

This [ambivalence] requires us to give centre stage to our mundane expe- riences of technology, and to all the contradictions and tensions involved:

technology is good and bad; it is enabling and it is oppressive; it works and it does not; and, as just part of all this, it does and does not have poli- tics. These tensions are a significant manifestation of the competing discourses to which our experience of technology is subject, and within which we make sense of them (p. 443).

Lessig’s point that “code codifies values and yet, oddly, most people speak as if code were just a question of engineering” (2006, p. 78) would appear to be incontestable. According to Lessig, if we take the view that code is law, we should opt for transparency in the regulation of cyberspace by means of open code. His concern is that “liberty will not take care of itself", which means that we should be aware of the values and norms embedded in software architectures and select those that guarantee a freer society (Lessig, 1999, p. 58). We should ask how code regulates, who the code writers are, and who controls the code makers (Lessig, 1999, p. 60; see also Lessig, 2006, p. 207). Lessig’s viewpoint highlights the embeddedness of values in technological decision-making.

Hence, one should ask questions related to power in and over code, i.e. questions related to transparency and openness in open source projects and IT governance. To who is participation open? Who are the developers and users?

Who controls the selection of software tools, and how is their use and imple- mentation regulated?

In order to understand where these general social implications come from, and how they resonate with recent developments, a necessary step is to examine some historical origins of the open development model.

2.2 Historical roots of the open development model

The purpose of this section is to provide a historical background for my research questions concerning the changing open development model. By understanding earlier events and different ways of organizing software production, we can gain better understanding into the current myriad forms and communities related to the development of open source.

The open development model of software production is intrinsically bound up with the changes in the nature of the computer. In the 1930s and 1940s the computer was the size of a room (mainframe computers and time sharing sys-

(25)

8

tems), and hence very different from today’s desktop computer. The larger dissemination of the PC to the masses has been linked to the maturation of the Internet and the World Wide Web (Ceruzzi, 1998). Various historical accounts and interpretations of origins of the open development model have been pre- sented (Ceruzzi, 1998, Aspray & Ceruzzi, 2008; Schwarz and Takhteyev, 2010). I have taken up from literature the historical events that are central to the argumentation put forward in this book. While Ceruzzi (2008) aptly reminds,

“One should resist any attempt to find a single point of origin of computer networks” (p.10), it should be kept in mind that the following historical narra- tive describes only some exemplary events that contributed to the emergence of the open development model, as it is known to date. The account, therefore, is not intended as any comprehensive history of open source software. Indeed, things were happening at the same time in many different places, and account- ing for all contributing events would be a task in its own right. The purpose of the following phasing is to draw attention to the change in the object of production and related developer “community”. Hence, the main point is that when we move from a product that it developed and used by the programmers them- selves, user-developers, to a product that is targeted for end-user use, the

“community” constellation and dynamics change: the “developer” and “user”

and no longer necessarily the same person. This leads to the constitutive problem in software design, namely, how to take into consideration to needs and requirements of the end user (e.g. Friedman, 1989, pp. 189).

I what follows, I will describe events from four different decades⁶, that have contributed to the emergence of the open development model. These events were identified on the basis of the form of organization and constellation of developers, the changing role of users, and the nature of the product developed.

What unites the three first three decades is that developers (primarily researchers or academics) were simultaneously users, and the products were initially built for own use. The fourth phase marks a movement from developing for own use to developing products targeted for end-users.

The phases and events are:

1. The development of the ARPANET in the US military force’s scientific networks, and the development of UNIX-operating system in the local scientific networks of Berkeley University by researcher-developers in the 1960’s and 70’s.

6 This phasing is adapted and slightly modified from Miettinen, Toikka, Tuunainen, Lehenkari, and Freeman (2006).

(26)

9 2. The institutionalization of open source: GPL and the Free Software

Foundation in the 1980s by Richard Stallman, a software “hacker”, researcher and idealist.

3. The establishment of open source software production communities and the Open Source Initiative (OSI) in the 1990s–Linux as the

paradigmatic example of Internet-mediated open development.

4. The emergence of hybrids of open source communities and

corporations in late 1990s and 2000s. OpenOffice.org and GNOME are examples of such hybrids.

2.2.1 ARPANET and UNIX: open development in local scientific networks in 1960–1970

Some of the early roots of the open development model can be traced back to the development of “ARPANET” and the “UNIX operating system” around the end of the 1960s. Both events are important to briefly describe because: 1) ARPANET is generally considered the predecessor of the Internet 2) The UNIX operating system underpins almost every system on the Internet (Ceruzzi, 2008, pp. 9-43, Ceruzzi, 1998, pp. 281–306).

The ARPANET grew out of a project funded by the U.S. Department of De- fense's Advanced Research Projects Agency, ARPA, (later “DARPA”) in the mid 1960’s (Ceruzzi, 2008, p. 9). Lawrence “Larry” Roberts, director of AR- PA’s Information Processing Techniques, learned about a new communication method for passing information in a network. The new packet switching computer network connected important research organizations in the United States.

Although access to ARPANET was limited even on university campuses at first, it later became more widely accessible within the scientific community.

Further, ARPANET-inspired similar packet switching networks were established in U.S. Universities thereafter, BITNET and USENET to name a few (Ceruzzi, 2008, pp. 9–43). In 1983 the ARPANET network switched this tech- nique to a set of protocols called the ”Transmission-Control Protocol/Internet Protocol (TCP/IP), fundamental to the working of the Internet, as we know it to date (Ceruzzi, 2008, p. 11). In essence, the people contributing to the development of ARPANET were academics (developer-users), who were thus able communicate and collaborate with each other from distance.

Also an important event in the history of the open development model was the development of the UNIX operating system. The key developers of UNIX were two American researchers from the AT&T Bell Labs, Dennis Richie and Ken Thompson. Later, researchers from the Universities of Berkeley, Carnegie- Mellon, and MIT also participated in its development (McKusic, 1999, p. 35).

In 1956 the US government prohibited AT &T from expanding its activity

(27)

10

outside the telecommunications business, which meant that UNIX could not be commercialized (Moody, 2001; Weber, 2004). AT & T’s strategy was to license UNIX and its source code to universities at a nominal fee of 100 USD. This way it could be used for educational purposes. Students could look at the source code and learn how it works. In 1975 UNIX was used in over 40 US Universi- ties and research laboratories (Weber, 2004, p. 29).

The development of UNIX contributed to a new conception of software development by enabling the incremental building of software so that many could contribute simultaneously (Ceruzzi, 1998, p. 332)⁷. UNIX acquired both technical and social characteristics that set it part from other programming systems, and contributed to the understanding Linux development (Ceruzzi, 1998, 332–

333). The benefits of UNIX and the C-programming language were for example transferability⁸, the simplicity of the software architecture, and flexibility–the combinability of different programs (Raymond, 2000, p. 23). The researcher- developer networks, both in universities and in companies, crossed organiza- tional boundaries, thus quickening the pace of development and changing the model of development. The Berkeley Software Distribution (BSD) became the largest and most efficient distribution channel for UNIX (Moody, 2001). Bill Joy, the co-founder of Sun Microsystems⁹, was among those involved in the BSD network (Ceruzzi, 1998; Raymond, 1999; McKusic, 1999; Moody, 2001).

Although UNIX was initially written by and for researchers, it also gained wider use (Ceruzzi, 1998, p. 284). The inclusion of Bil Joy’s TCP/IP stack in to the BSD UNIX was instrumental in the democratizing of internetworking because it helped transform the restricted ARPANET to the wide-open Internet (Ceruzzi, 1998, p. 284).

2.2.2 Institutionalization of open source: the General Public License and the software foundation in the 1980s

A key enabling mechanism in the institutionalization of open development was the “Copyleft”- principles and the General Public License (GPL). These were pioneered and mostly written by Richard Stallman, a well-known Free Software advocate. Stallman formulated them to keep his own projects, “Gnu Emacs”

text editor and the “Gnu Hurd” operating system, developed originally as alter- natives to commercial development of UNIX in 1983, free for everybody.

Although open source software licenses come in many forms (Siltala, 2003),

7 See also: http://www.faqs.org/docs/artu/ch01s06.html.

8 Transferability means that a computer program is easily transferred from one operating system to another.

9 The sponsoring firm of the OpenOffice.org studied in this book.

(28)

11 what distinguishes them from proprietary ones is that they are based on an entirely different conception of intellectual property. The source code is generally kept open so that anyone can view, modify and in some cases, distribute their modifications to others. However, what many refer as “public domain software” is a legal term for software that is not copyrighted. Stallman empha- sizes that this should not used for talking about free software.¹⁰ Hence, Copy- lefted software is free software whose distribution terms warrant that all copies of all versions carry similar distribution terms. The objective is to guarantee that the software programs remain freely available, distributable and modifiable.

The licence also restricts software from ending up in commercial ownership (Stallman, 1984).

Copyleft refers to “a general method for making a program (or other work) free, and requiring all modified and extended versions of the program to be free as well” (http://www.fsf.org/).

Stallman defined free software in 1984:

Free software is a matter of the users' freedom to run, copy, distribute, study, change and improve the software. More precisely, it refers to four kinds of freedom, for the users of the software: The freedom to run the program, for any purpose (freedom 0). The freedom to study how the program works, and adapt it to your needs (freedom 1). Access to the source code is a precondition for this. The freedom to redistribute copies so you can help your neighbor (freedom 2). The freedom to improve the program, and release your improvements (and modified versions in general) to the public, so that the whole community benefits (freedom 3).

Access to the source code is a precondition for this. A program is free software if users have all of these freedoms (Stallman, 1984, p. 8).

For Stallman and other advocates of the free software movement, freedom is attained through being able to use a program for any purpose: to study, redistribute, improve and re-release improved code. Hence, “free” did not refer to gratis but liberty:

As our society grows more dependent on computers, the software we run is of critical importance to securing the future of a free society. Free software is about having control over the technology we use in our homes, schools and businesses, where computers work for our individual

10 http://www.gnu.org/philosophy/categories.html.

(29)

12

and communal benefit, not for proprietary software companies or gov- ernments who might seek to restrict and monitor us. (http://www.fsf.org/) The Free Software Foundation (FSF) was established in 1985 for the purpose of pursuing the freedom of code and freedom of information. The FSF advocates free software adoption and campaigns against proprietary software.¹¹

2.2.3 The establishment of open source software production com- munities in the 1990s

While important institutional foundations of the open development model were established earlier, this phase is marked by the emergence of globally dispersed open source production communities powered by Internet technologies. Be- cause the Linux community is used as a reference point and paradigmatic example of such development (e.g. Benkler, 2003; von Hippel & Von Krogh, 2003; von Hippel, 2005; Weber, 2004), its trajectory and unique ways of organizing development are briefly recalled. According to Ceruzzi (1998), the wide use of UNIX in universities and other non-commercial locations contributed to the emergence of Linux, since it is through UNIX that developers started using what they were developing (p. 332).

In 1991 a Finnish computer science student and hobbyist, Linus Torvalds experienced frustration with the MS DOS operating system he was using, because it was unreliable and far too expensive. He gained acquaintance with UNIX and the C programming language during his university studies. He had also encountered Minix, a version of UNIX, developed for educational purposes by the American professor Andrew S. Tannembaum. Although the Minix code was available for view, it was not freely modifiable. All modifications to the Minix source code had to be approved by Tannembaum. The problem of not being able to modify the source, and the fact that it was hard to contact Helsinki University UNIX machines with Minix, led Torvalds to develop a necessary and more reliable operating system (Moody, 2001, p. 31). In the beginning, Torvalds was the only one developing the code (Torvalds & Diamond, 2001, p.

102).

Torvalds followed Internet newsgroups and posted an announcement to a UseNet newsgroup “comp.os.minix” thus revealing his intention to develop an operating system. Developers from around the globe answered his post and expressed their willingness to test his code. A couple weeks later the first version of Linux was downloadable from the Helsinki University server. Torvalds collected the names and email addresses of all those willing to test the system

11 http://www.fsf.org/.

(30)

13 and announced that they could start straight away (Torvalds & Diamond, 2001, p. 102; Moody, 2001, p. 35). The mailing list became an important collaboration tool in the Linux project (see Torvalds & Diamond, 2001). The project started out from Torvald’s personal need for an operating system and quickly grew into a global volunteer Internet-based “virtual” project. At this stage, the development of Linux changed from a one-man project to a community effort.

The people engaged in the project consisted of university students, professors and self-taught hobbyist programmers–“hackers”–, thus forming a collaborative network of user-developers, developers who were also users but not all necessarily academics. The driving force of the community was competition over who would develop the best code and thus succeed in getting his/her piece of software integrated into the “core”. At first all code modifications had to be approved by Torvalds himself, the leader and decision maker in the project (Moon & Sproull, 2002, p. 390).

As the project grew, the code had to be modularized. According to Torvalds and Diamond (2001, p.108), modularity–the formation of a system from independent but combinable parts–was the crucial factor for the development of Linux. However, proprietary software developer-companies like SAP¹² and Microsoft¹³ also utilize modularity in their outside user and developer communities. In this sense the boundaries between the closed model and the open model appear fuzzy.

The Internet and modularity together can be seen as the enabling conditions for the simultaneous development of software irrespective of geographic location or time zone (Moon & Sproull, 2000). Modularity also affected the division of labour: for every module a person in charge was appointed who was respon- sible for the module’s internal decision making. Gradually Torvalds gave a few chosen and trusted coders more power in decision-making (Iannaci, 2005, p. 16). It has been suggested that the project was organized into two “circles”

(Moon & Sproull, 2000). The inner circle consisted of Linus Torvalds and his trusted module maintainers who “filtered out” code contributions. The outer circle consisted of a large crowd of programmers who reported bugs and made suggestions for improvements. Torvalds did not meet the other developers face to face for many years (Torvalds & Diamond, 2001, p. 151). The Internet is still the most important development tool and forum, and the quality of code would seem to be the main criteria in building trust (Torvalds & Diamond, 2001). If

12 The acronym SAP stands for “Systems, Applications, and Products in Data Processing”. SAP is a professional social network community of SAP customers, partners, employees and experts (http://www.sdn.sap.com/irj/scn).

13 Microsoft technical communities provide opportunities to interact with Microsoft employees, experts, and your peers in order to share knowledge and news about Microsoft products and related technologies. (http://www.microsoft.com/communities/default.mspx).

(31)

14

the quality of code meets the standards of Torvalds or his “lieutenants” (the module maintainers), the code committer is named in the Linux Credits File as payment for a job well done (Moody, 2001, p. 14). The person in charge of a module communicates with the external group of module maintainers. This is how a more complicated hierarchy was established in the project (Moon &

Sproull, 2002).

Linus Torvalds also joined the Open Source Initiative (OSI)¹⁴ founded in 1998 by Eric Raymond and Bruce Perens. The purpose of the OSI is to actively participate in Open Source community building and education, and to function as public advocacy for promoting awareness and the importance of non- proprietary software. The term “open source'” was invented to promote the new rhetoric of pragmatism and market-friendliness that Raymond had been developing.

Open source is a development method for software that harnesses the power of distributed peer review and transparency of process. The prom- ise of open source is better quality, higher reliability, more flexibility, lower cost, and an end to predatory vendor lock-in.¹⁵

“Sourceforge” hosts around 318, 008¹⁶ open source code repositories while

“github” hosts 3, 098, 237 different repositories¹⁷.

2.2.4 The changing relationship between communities and firms in the 2000s

In the early years of Linux development its source code was mainly used as a platform for the further development of the code itself. When Linux started to turn into a viable and competitive alternative to proprietary software, people characterized as “end-users”–users that were not able to write or modify code–

started using it. For such end-users, Linux was not a complex system of interacting source code modules and programming tools but a resource and tool in the office (Tuomi, 2001, p. 16). This new group of users contributed to the

14 Open source is defined by the OSI as follows: 1. Free Redistribution 2. Source Code 3. Derived Works 4. Integrity of The Author's Source Code 5. No Discrimination Against Persons or Groups 6. No Discrimination Against Fields of Endeavor 7. Distribution of License 8. License Must Not Be Specific to a Product .9. License Must Not Restrict Other Software 10. License Must Be Technology-Neutral (http://www.opensource.org/docs/osd).

15 See http://www.opensource.org/docs/osd.

16 See http://sourceforge.net. This was the situation in 2011.

17 See http://www.github.com. This was the situation in 2011.

(32)

15 emergence of two kinds of firm-community relationship, in which both parties are dependent on each other (Dahlander & Magnusson, 2005, p. 487).

The first type of community-firm relationship can be characterized as sym- biotic: both the firm and the community benefits from collaboration (Dahlander

& Magnusson, 2005, p 487). The aim of the firm’s business model is to package Linux to better suit the needs of the end user by adding a desktop environment and window system, as well as installation and user manuals to the kernel.

Linux distributions (e.g. Red Hat and SUSE) are good examples of firms that are dependent on the open development model. The distributors also allow their developers to participate in the development of Linux. Although these packages are often distributed free of charge, a service provision fee is collected. Nowa- days corporate-paid employees such as programmers and marketing- and sales personnel or distribution firms also participate in the development of Linux.

The open development model of producing software has firmly established itself as part of the software industry, and as a new form of software practice.

The current phase of the open development model is characterized by the simultaneous competitiveness and intertwining of two very different methods of software production: open and proprietary. Further, this phase is also marked by the shift from developing programmer software to taking into account end-user needs either by packaging Linux as exemplified above or by starting to produce end-user software, as will be discussed next.

The second type of firm-community relationship emerged in 2000 when companies initiated open source projects by freeing in-house built code to volunteers. So-called hybrid open source projects, in which volunteers and firms collaborate, started to emerge. The corporate has representatives in the governing bodies and its employed staff work as module maintainers and project leaders, and contribute substantially to the core code construction (Shah, 2007; Siltala, Freeman, & Miettinen, 2007).

Siltala et al. (2007, p. 23) identify two types of firm-hybrid communities:

firm-driven, “top down”, in which was born out of firm-owned code (e.g. Ope- nOffice.org), and community-driven “bottom-up” originating from community developed free software but later attracting the interest of firms (e.g. GNOME).

The first firm-driven hybrid before OpenOffice.org emerged when Netscape released its Web browser¹⁸ Netscape Navigator source code to open source

18 “A web browser is a software application for retrieving, presenting, and traversing information resources on the World Wide Web. An information resource is identified by a Uniform Resource Identifier (URI) and may be a web page, image, video, or other piece of content. Hyperlinks present in resources enable users easily to navigate their browsers to related resources. A web browser can also be defined as an application software or program designed to enable users to access, retrieve and view documents and other resources on the Internet.

http://en.wikipedia.org/wiki/Webbrowser.

(33)

16

developers in 1998. The Mozilla project was born, and with it the Firefox web browser. Netscape’s role in the commercialization of the Internet cannot be underestimated (Ceruzzi, 2008). The hybrid community-driven GNOME was established in 2000 with the birth of the GNOME Foundation, which is a steer- ing committee and legal branch of the GNOME project. Novell, Inc. and Red Hat, Inc together form a majority on the Foundation’s eleven-member Board of Directors (Siltala et. al., 2007, p. 13). Sun Microsystems followed Netscape’s example and released the code of its proprietary StarOffice¹⁹ In 2000 Sun Mi- crosystems aimed at building a volunteer community around the existing code base and named it the “OpenOffice.org project”.²⁰ Sun (later, Oracle) was the main sponsor and developer of the OpenOffice.org code, thus remaining in control of overall development. Here “hybrid community” refers to the firm having representatives in the sub-projects and governing body of OpenOf- fice.org. The JCA (Joint Copyright Agreement)²¹ and the LGPL license used in the project allowed the company to use the OpenOffice.org code in developing its proprietary StarOffice.

2.3 Research task and research problems

The research task of this study emerged from the historical developments in the open development model. The movement from “traditional” open source hacker projects like Linux to hybrid forms of open source such as the OpenOffice.org was characterized by change in the object of development. In the Linux community, the object was an operating system developed for and by other programming skilled user-developers. However, in the hybrid OpenOffice.org project initiated by Sun Microsystems the intent was to develop an office application intended for end-users by establishing a volunteer community around an existing code-base. This inevitably changes the community and the discourse used to describe the community. The dominating discourse on open source mainly characterizes programmer-to-programmer projects, shown in the triangle on the left in Figure 1. This study seeks to contribute to filling a gap in the existing research by investigating, with a view to identifying and explaining, the changing community and discourse of the hybrid open source community.

19 Perhaps the software most commonly known to the average computer user is Office software (Office productivity). It typically includes functions such as word processing, spreadsheets, presentations, drawings, web publishing, email, scheduling, and database applications”. Other major browsers are the Internet Explorer, Google Chrome, Safari, Opera and SeaMonkey.

“(http://en.wikipedia.org/wiki/World_Wide_Web)

20 A historical account of Sun Microsystems is given in chapter 5.

21 See Appendix 1.

(34)

17

Figure 1. The research task: the challenge of understanding the discourse of community and its membership in the transition from hacker to end-user oriented communities. For the concept of object-oriented mediated activity system, see En- geström (1987, p. 78).

On the basis of the transition depicted in Figure 1, I pose the following general research questions: how is the structure and membership constellation of the community, specifically the relation between developers and users linguistically constructed in hybrid open development? What characterizes Internet-mediated virtual communities and how can they be defined? How do they differ from hierarchical forms of knowledge production on one hand and from traditional volunteer communities on the other?

These general questions will be addressed through the following four more specific research questions:

1. How is the hybrid OpenOffice.org community viewed by OpenOffice.org Groupware volunteer contributors?

Do they find the combination of open source principles and business activity compatible?

One can presume that in a firm-community hybrid collaborative project like the OpenOffice.org different conflicting interests will come together and collide.

The core difference between open source and closed in-house software development is in the way intellectual property (in the form of software code) is conceived and defined. While proprietary software businesses hold to the source code as a trade secret, open source licenses assure the user the freedom to access, modify and re-distribute software (e.g. Stallman, 2003). Thus, combining different underlying principles of intellectual property and cultures of soft-

Discourse on community and membership:

Hacker ethic; the Bazaar model and user empowerment

Hybrid Discourse: ?

Community:

“Hackers”, user-developers

Community: ? Object/product:

programmer software tools

Object/product:

end-user software

Constructing a Community : Myths and Realities of the Open Development Model

Stephanie Freeman